Modelling & Dynamic Analysis of

Wind Turbine Blades

by
Mr. Viswa Teja Vanapalli
(Roll No: 213ME1397)

Under the guidance of

Dr. (Prof.) J. Srinivas

Department of Mechanical Engineering,

National Institute of Technology, Rourkela,
Rourkela – 769008, Odisha, India.
Modelling & Dynamic Analysis of
Wind Turbine Blades

A thesis submitted in partial fulfilment of the

requirements for the degree of

Master of Technology
in
Machine Design and Analysis
by
Mr. Viswa Teja Vanapalli
(Roll No: 213ME1397)

Under the guidance of

Dr. (Prof.) J. Srinivas

Department of Mechanical Engineering,

National Institute of Technology, Rourkela,
Rourkela – 769008, Odisha, India.
 NATIONAL INSTITUTE OF TECHNOLOGY, ROURKELA

CERTIFICATE
This is to certify that the thesis entitled, "Modelling &Dynamic Analysis of Wind

Turbine Blades" submitted by Mr. Viswa Teja Vanapalli (213ME1397), submitted

to the National Institute of Technology, Rourkela for the award of Master of

Technology in Mechanical Engineering with the specialization of “Machine

Design & Analysis”, is a record of bonafide research work carried out by him in

the Department of Mechanical Engineering, under my supervision and guidance. I

believe that this thesis fulfils part of the requirements for the award of the degree of

Master of Technology. The results embodied in this thesis have not been submitted

for the award of any other degree elsewhere.

Dr. (Prof.) J. Srinivas

Dept. of Mechanical Engineering
National Institute of Technology,
Place: Rourkela Rourkela-769008, Odisha,
Date: India.
 ACKNOWLEDGEMENT

I wish to express my deep sense of gratitude to Dr. J. Srinivas, Professor,

Mechanical Engineering Department, for his wholehearted co-operation, unfailing

inspiration and valuable guidance.

It is a privilege to express my deepest gratitude to Dr. S. K. Sarangi, Director and

Professor Dr. S.S. Mahapatra, the Head of Department for extending the utmost

support and cooperation to provide all the provisions & facilities for the successful

completion of the project. I am extremely grateful to Mr. Prabhu Lakshmanan for

his untimely support and his encouragement. I also wish to thank Ms Varalakshmi,

Satish Teli & Ashish Gurjar for their suggestions throughout the project and their

helpful company.

I also sincerely thank, the supporting staff for their sustained help in the successful

completion of my project and all those who contributed directly or indirectly in

successfully carrying out this work.

V. VISWA TEJA

Rourkela, May 2015

 ABSTRACT
Wind turbine blade is one of the important components requiring more attention at the design

stage. These blades are made up of fibrous material and sometimes hollow composite web sections

may be employed in its construction. The main focus in its design is to achieve a desired strength

to withstand various loads as per the power requirements. In view of this, modelling & dynamic

analysis of blades is an essential requirement not only to avoid resonant vibrations but also to

understand the stability of operation during various operating conditions. In this line, present work

focuses on dynamics of horizontal- axis wind turbine blade with NACA 63415 profiles subjected

to aerodynamic, centrifugal and gravity loads. The basic geometrical parameters of the blade are

designed by using blade element moment method & the blade is modelled as a ruled 3D surface.

By computing element wise cross sectional details from 3D model, a 1D beam finite element

modelling is developed and modal characteristics are obtained. These are further validated with

3D finite element model result. Then, the effects of the tip speed ratio and rotational speed on the

natural frequencies are studied. Further, finite element model have 10 elements with 6 degrees of

freedom per node is used to obtain the dynamic response of the blade subjected to various loading

conditions. The effect of aerodynamic, centrifugal and gravity parameters on the frequency

response, edge-wise and flap-wise beating at the tip of blade are studied. The iterative programs

developed in the work helps in testing the blade frequencies & response at any operating

conditions. In order to estimate the stability during rotation, the frequency responses are obtained

from response histories at the blade tip, by using Fast Fourier Transform algorithm. It is found that

the centrifugal loads have profound effect on the frequency responses compared with other loads.

The dynamics of long blade when rotating at varying wind conditions (aerodynamic loads) is

influenced by gravity loads also.

 INDEX
NOMENCLATURE

LIST OF FIGURES

LIST OF TABLES

CHAPTER-1 INTRODUCTION ............................................................................................ 1

1.1 Wind Turbine Blades ...................................................................................................... 2

1.2 Literature Review............................................................................................................ 4

1.3 Scope and Objectives .................................................................................................... 10

1.4 Thesis Organization ...................................................................................................... 11

CHAPTER-2 MATHEMATICAL MODELLING ............................................................. 12

2.1 Equations of Motion of a Rotating Wind Turbine Blade .............................................. 12

2.2 Approximate Solution approach ................................................................................... 14

2.3 Loads Acting on Wind Turbine Blade .......................................................................... 15

2.3.1 Aerodynamic Loads ............................................................................................... 16

2.3.2 Centrifugal & Gravitational Loads: ....................................................................... 17

2.3.3 Gyroscopic Load.................................................................................................... 17

CHAPTER-3 FINITE ELEMENT MODEL OF THE BLADE ........................................ 18

3.1 Blade Discretization ...................................................................................................... 18

3.2 3D Modelling of the Blade ........................................................................................... 21

i
CHAPTER-4 RESULTS & DISCUSSION.......................................................................... 24

4.1 Blade Data Selection Procedure.................................................................................... 24

4.2 Input Parameters ........................................................................................................... 28

4.3 1-D Beam Finite Element Analysis Outputs ................................................................. 29

4.3 Modal Characteristics of the blade from ANSYS 15.0................................................. 30

4.4 Effect of Rotational Speed on Natural Frequencies ...................................................... 33

4.5 Effect of tip speed ratio on the chord & Twist distributions of the blade ..................... 36

4.6 Effect of tip speed ratio on the modal characteristics of the blade ............................... 37

4.7 Dynamic response Analysis .......................................................................................... 38

CHAPTER-5 CONCLUSION............................................................................................... 45

5.1 Summary ....................................................................................................................... 45

5.2 Future Scope ................................................................................................................. 46

REFERENCES ....................................................................................................................... 47

APPENDIX I .......................................................................................................................... 51

APPENDIX II ...................................................................................................................... 55

APPENDIX III ....................................................................................................................... 58

APPENDIX IV ....................................................................................................................... 60

ii
 NOMENCLATURE
Ae Area of Blade Element
a axial induction factor

a’ angular induction factor

B Number of blades

CL lift coefficient

CD drag coeffient

c aerofoil chord length

Cp Power Coefficient

dFt Thrust force

dFd Drag force

dM Pitching Moment

Ie Moment of Inertia of Blade element

N Number of Blade elements

λ Tip speed ratio

λr local tip speed ratio

Ω blade rotational speed

φs setting angle

σ’ Local Solidity

ρ density of the Glass fiber reinforced plastic material

ρa density of the air at 15°C standard conditions

iii
 LIST OF FIGURES

Fig 1.1 Airfoil profile Nomenclature ................................................................................... 3

Fig 2.1 Blade configuration ............................................................................................... 12

Fig 2.2 Representation of Blade elements and blade geometry ......................................... 15

Fig 3.1 Discretization of the blade .................................................................................... 18

Fig 3.2 Euler configuration of the blade ............................................................................ 20

Fig 3.3 Solid model of the twisted blade ........................................................................... 22

Fig 3.4 Imported solid model ............................................................................................. 22

Fig 4.1 Aerodynamic profile of NACA 63-415 airfoil ...................................................... 24

Fig 4.2 Flowchart to determine blade’s geometric parameters .......................................... 25

Fig 4.3 Chord distribution of the blade .............................................................................. 26

Fig 4.4 Plots of Lift, Drag and Moment Coefficients of NACA 63415 ............................ 27

Fig 4.5 Twist distribution of the blade ............................................................................... 28

Fig 4.6 Mapped Mesh of the blade .................................................................................... 31

Fig 4.7 Mode shapes observed in Modal analysis conducted by ANSYS ......................... 33

Fig 4.8 Variation of Flap wise bending mode with respect to rotational speed ................. 34

Fig 4.9 Variation of Torsional bending mode with respect to rotational speed ................. 34

Fig 4.10 Variation of edgewise bending mode with respect to rotational speed ................. 35

Fig 4.11 Chord distribution variation for different taper ratios. .......................................... 36

Fig 4.12 Twist distribution variation for different taper ratios ............................................ 36

Fig 4.13 Modal Variation for different taper ratios ............................................................. 37

Fig 4.14 Variation of flap wise, edgewise and torsional modes for different taper ratios. .. 38

Fig 4.15 FFT spectrum of blade dynamic response. ............................................................ 39

Fig 4.16 Edge-wise beating at blade end ............................................................................. 40

iv
Fig 4.17 Flap-wise beating at the blade end ........................................................................ 40

Fig 4.18 Edge-wise beating at blade end subjected to only aerodynamic loads .................. 41

Fig 4.19 Flap-wise Beating at the blade end subjected to only aerodynamic loads ............ 41

Fig 4.20 FFT spectrum of blade dynamic response subjected to only centrifugal loads. .... 42

Fig 4.21 Edge-wise beating at blade end subjected to only centrifugal loads ..................... 42

Fig 4.22 FFT spectrum of blade dynamic response subjected to only centrifugal loads. .... 43

Fig 4.23 Flap-wise beating at blade end subjected to only centrifugal loads ...................... 44

Fig A1 Aerodynamic profile of NACA 4412 profile ........................................................ 52

Fig A2 Profile of NACA 63-415 airfoil drawn in MATLAB............................................ 53

v
 LIST OF TABLES

Table 4.1 Properties of the blade ....................................................................................... 28

Table 4.2 Inertial Properties of the Blade at tip-speed ratio λ=6 ....................................... 29

Table 4.4 Modal Characteristics of the Blade .................................................................... 31

Table A1 Geometrical Properties of Blade Elements obtained from initial program for Tip

speed ratio λ=5. .................................................................................................. 55

Table A2 Geometrical Properties of Blade Elements obtained from initial program for Tip

speed ratio λ=6. .................................................................................................. 56

Table A3 Geometrical Properties of Blade Elements obtained from initial program for Tip

speed ratio λ=7. .................................................................................................. 57

vi
 Chapter-1

Introduction
Wind turbines can be classified into Horizontal Axis Wind Turbines (HAWT), and Vertical

Axis Wind Turbines (VAWT). Horizontal type turbines have the blades rotating in a plane

which is perpendicular to the axis of rotation. The HAWTs are most widely used type of wind

turbines and come in varied sizes and shapes. The primary type of force acting on the blades

of HAWT are the drag forces. The Horizontal type are commercially applicable type due to its

varied sizes and storage capacities. In Vertical Axis Wind Turbine blades, the rotation axis

coincides with the axis of generation of power. Advantages of Vertical Axis wind turbine is

the placement of generator at the base of the turbine. But these type are nearer to ground which

makes it difficult to capture the power at higher altitudes. While the HAWTs require yaw

mechanism to orient themselves in the direction of wind, the VAWTs do not have such

problem. But disadvantage with VAWTs is the low starting torque, lesser installation height

and dynamic stability issues.

Any Horizontal axis wind turbine contains five major components: Rotor Blades, Rotor Hub,

Nacelle, Yaw system and Tower. It is the shape of Blade which decides the tapping of wind

energy and conversion into Kinetic energy of the blades. These blades have in general

aerodynamic profiles and are constructed with materials having high strength to weight ratio.

The tip speed ratio, number of blades and profile of the blades create huge difference in

generation of wind power. The hub connects the rotor blades with the generator shaft. The

nacelle houses a gearbox and generator to tap the power obtained at the shaft into electrical

energy. A yaw mechanism is provided to the HAWT in order to orient itself to the wind

direction. The sensitivity of the yaw mechanism plays important role in making the blade orient

1
itself towards the wind and improving efficiency of the turbine. The Tower provides the

necessary altitude for the blades to overcome the obstacles that might cause turbulence in wind

at smaller heights.

1.1 Wind Turbine Blades

Horizontal axis wind turbine blades are subjected to various type of excitations. Wind turbine

blades has an aerofoil cross section with taper & twist incorporated to generate more output

energy. During the rotation, the blade is subjected to centrifugal forces, and other aerodynamic

effects as well as the gyroscopic coupling effect. Several earlier works studied these dynamic

analysis problems of wind turbine blades. The modelling of the blade is carried out as a

cantilever beam with unsymmetrical and variable cross section. Routine approach for study of

such blades is finite element modelling where the beam is discretized into several elements,

each having a defined area and Moment of Inertia. In practice for a lengthy wind turbine blade

of unsymmetrical section, the element stiffness & mass matrices contain in total four Moments

of Inertia, Ixx, Iyy, Izz & Iyz.

Airfoils are common profiles for wind turbine blades. The blade contour must utilize the

aerodynamic considerations while the material provides the stiffness and strength to the blade.

An airfoil is defined by a number of terms as shown in Figure 1.1. The leading edge is where

the wind starts to hit the airfoil and trailing edge is where the air leaves the airfoil surface. The

upper surface and lower surface of airfoil have a mean camber line midway between the two

lines. The chord line connects the leading edge and trailing edge and the distance is designated

by c. Camber is the maximum distance between the chord line and mean camber line. The angle

of attack α is the angle at which the wind strikes the airfoil with relative wind velocity Vrel. The

thickness of the airfoil is distance between the upper & lower surface of the airfoil. The chord

length, thickness and camber affect the aerodynamic performance of the airfoil.

2
Angle of attack Mean Camber
Line

α Trailing edge
Vrel

Leading edge Chord line

Fig 1.1 Airfoil profile Nomenclature

There are three widely wind turbine profiles which are NACA, LS & LM profile standards.

Out of these, NACA airfoils have been widely accepted even in commercial applications due

to their high power coefficients and lesser drag profiles. NACA stands for “National Advisory

Committee for Aeronautics”. In Horizontal axis wind turbines, the NACA profiles are used to

generate the profile shape of the cross section of the blade by using a set of camber line

equations to generate points on the upper and lower surface of the airfoil. The Shape of NACA

airfoil is to be decoded from a series of digits which when substituted in the Equations gives

the coordinates of the airfoil. There are different NACA series available for various aerospace,

wind energy, rocketry applications out of which NACA 4 digit series and NACA 6 digit series

provide better performance when used for generation of wind power.

3
1.2 Literature Review

In the area of wind turbine blade design and analysis several works are available in literature.

These are summarized in this section.

Aerodynamic Design & Blade Optimization of Wind Turbine Blades:

Designing a blade by using well known Blade Element Momentum (BEM) Theory is a

fascination to many researchers who are interested to get better geometrical parameters to

maximize the power output of the blades.

National Advisory Committee for Aeronautics (NACA) developed a series of standard airfoil

shapes which are prominent in usage as the profiles of the blades. Using the airfoil profile,

knowing the angle of attack of the wind and the proposed length of the blade, researchers such

as Tenguria et al. [1] have used BEM theory to optimize the coefficient of power, lift and drag

characteristics with various tip speed ratio, lift and drag coefficients. It is observed that at a

power coefficient 0.46 and lift to drag coefficient is 124.47, the power absorption is maximum.

John McCosker [2], developed an optimized code for a discretized 9 element wind turbine

blade having a length of 0.95 m. He obtained optimal speed ratio, angle of wind, the pitch angle

and relative chord lengths for each element. After convergence, the power extracted from wind

is found to be 0.81 KW. The airfoil shape is varied from NACA 4412 to NACA 23012 where

the angle of attack of the blade to get maximum glide ratio is found different for each profile.

Cost reduction of generation of wind power is critical and it can be possible by suitable

structural design changes and use of composite materials to lead greater profits. This is

discussed by Anjali et al [3], in which they described an ant colony optimization method by

varying the composite material of the blade and conducting stress analysis. Ultimately, he

4
obtained the optimized values of chord length and blade twist angles and it is found that Kevlar

149 material has highest natural frequency and less deflection compared to others.

Thumthae and Chitsomboon [4] investigated an untwisted blade for obtaining optimal pitch to

get maximum power for wind turbine using stead flow wind conditions. Research is also done

using Genetic algorithms to obtain the optimization of chord and twist angle by Juan Mendez

and Greiner [5], for a test turbine having 19m rotor diameter and 100 KW nominal power.

Similar works have been done [6-9] where observations state that optimizing the blade shape

parameters is mostly to maximize the output power generation of the blade.

Ingram [8] derived equations relating to the Blade Element Theory in order to find the axial

force, Lift and drag characteristics, considering the tip loss correction to calculate the rotor

performance. An iterative procedure is used to refine the obtained results.

Vasjaliya et al [9], evaluated multidisciplinary optimization process to minimize cost, weight

and maximize power output by considering a fluid structure interaction and structural

robustness to enhance performance of a QBlade / XFoil airfoil blade. In process of

optimization, structure and design variables are taken as input parameters and a number of

DOEs were created and solved.

Sedaghat et al. [10] used BEM Theory to get a generalized quadratic equation on angular

induction factor with tip speed ratio, drag to lift coefficient. Then an optimal blade geometry

is obtained which is used to calculate the power performance at variable speeds. BEM theory

ensures quicker way to understand the off design power performance of blades moving at a

constant speed.

Using 3D modelling softwares such as CATIA, Solidworks, the design of wind turbine blade

can be seen much more realistically. These surface models can even be exported to analysis

5
packages to conduct further research analysis by posing suitable conditions. Scott Larwood et

al. [11] studied the design of a swept wind turbine blade. Parametric study to determine sweep

parameters using STAR7d scaled model showed that loads were most sensitive on the tip of

sweep.

Finite Element Analysis & Modal Analysis of Wind Turbine Blades:

Today, the method of finding the modes of vibration of a structure has become

widespread to assess the inherent properties of the structure. The significance of finding the

solution for these single degree of freedom systems is such that they can be used to analyse the

complex multiple degree of freedom systems as the latter can be decoupled into a system of

Single degree of freedom systems. [12-16] discusses the experimental and numerical

investigations done on performing Modal Analysis on complex wind turbine blades.

Gursel et al. [12] studied the vibration characteristics of rotor blades using approximation

method such as Rayleigh to calculate the natural frequency of each blade. They have validated

the results of vibration analysis by using Finite element analysis.

Larsen et al. [13] shown experiments on a LM 19m blade to investigate the mode shapes,

dominating deflections, and explained the difference or error in measurement between the

theoretically computed values and experimental results. It is observed that for non – dominating

deflection direction, the measured and computed mode shapes are found to be in good

agreement. A forced vibration damper is introduced in the experimental setup to check the

damping characteristics of the blade.

Liu et al [14] studied modal and harmonic analysis of circumferential force is on a 5MW S809

airfoil, wind turbine model to obtain the natural vibration characteristics to get first seven

orders of natural frequencies along with harmonic responses in different phases.

6
Allikas et al. [15] validated a full scale single layer layup small horizontal axis wind turbine

blade through experimental bending test and modal analysis. They have taken Glass fiber

reinforced composite plastics to get the stiffness and strength analysis acted by 7848 N load. A

difference if 16.8% occurred during load Case 6, damaging the blade due to value of obtained

stress being greater than yield strength of the skin element of blade.

Sami et al. [16] extracted fundamental flapwise and edgewise modal frequency of a 5KW

GFRP wind turbine blade by using 3d shell elements. It is to understand better the dynamic

behaviour that he conducted experiments using electrodynamic shaker system to predict the

resonant frequencies. He observed that flapwise frequencies are found to be in agreement to

each other while % of error is more in case of edgewise frequency.

Fangfang song [17] worked on optimization of design of the blade, having NACA 63415

profile and then modelled the surface model of blade using Solidworks software. Then the

finite element model is considered to find out the modal analysis of the blade. The excited

frequeny from wind speed of 10m/s is calculated as 7.16 Hz which is found to be more than

the fundamental frequency obtained from the modal analysis therefore no resonance will occur

when the blades run at rated wind speed.

Dynamic Analysis:

Jie et al [18], used a 38m blade having rated power 1500KW and conducted structural stress

and strain distribution analysis to understand the flapwise loads and vibration mode shapes.

Shell 99 elements are used to discretize the blade using Finite Element Analysis to validate the

result. The authors have conducted structural response characteristics [19] to study the

aerodynamic characteristics using CFD simulation and then formulating dynamic

characteristics of the blade. The BEM method predicted much higher aerodynamic loads than

7
CFD method. A maximum error of 4.86% is observed between FE model and calculated

frequencies.

Inoue et al. [20] studied the dynamic analysis of wind turbine blade by investigating its

fundamental vibration behaviour. Effect of gravitational and wind load on super harmonic

resonance was presented.

Lag wise dynamic characteristics of a wind turbine blade subjected to unsteady aerodynamic

forces was studied by Li et al. [21]. A mathematical model [22] is developed for describing

non-linear vibration of horizontal axis wind turbines which uses Kelvin Voigt Theory to

compute the decouple a set of coupled equations of motion representing a wind turbine blade

subjected to aerodynamic and gravitational loads. The expressions for static deformation, aero

elastic stability and dynamics of the blade are solved from the set of equations. The system

consists of a rotating blade with out of plane bending having in plane bend and torsion. Factors

such as coning angle, twist angle, eccentricity, mass centre, shear centre were included.

Li et al. [23] also conducted the dynamic response analysis for flap wise direction in case of

super harmonic response. Amplitude modulation equations are used to derive frequency

response. Effect of static displacement, perturbation frequency, dynamic displacements are

studied and results are compared with numerical solution. It is found that dynamic

displacements of flap wise are larger compared to axial displacements of the blade.

Hamdi et al. [24] independently presented forced vibration analysis of wind turbine blade

rotating at a constant angular velocity by taking into consideration aerodynamic, centrifugal,

gravity and gyroscopic loads. In this, both static and dynamic investigation studies have been

discussed with and without gyroscopic loads is done. The excitation force vector containing

harmonic components is incorporated in the equations of motion, and the differential equations

8
are solved by using Newmark method. Dynamic responses along with FFT responses are

obtained for blade for the first 25 seconds.

Karadag [25] studied effect of shear center on dynamic characteristics of the blade. Different

methods such as Reissner, Potential and finite element methods are used to calculate the natural

frequencies of the blade rotating at 0 rpm and 3500 rpm for the loads acting at the shear centres.

The bending and torsional modes are observed and variation is found for finite element method

and Reissner method. It is seen that shear center affects the natural frequencies and modes of

the rotating beams with complex geometry. Both thick beam and thin beam theories give results

in close comparison. Torsional frequencies show greater change with increase of rotation than

the bending frequencies.

Nymann et al. [26] described a formulation of 3D two node FE analysis in rotating frame of

reference. The effect of Elastic, geometric and torsional bending are shown to play significant

role in improving accuracy modelling of structures. Structural responses at middle of the blade

and blade tip is observed. The actuator forces seem to reduce material strains at all times.

Keerthana et al. [27] introduced a step wise procedure to develop the blade’s geometrical

properties by using optimization techniques and taking input parameters as the tip speed ratio,

wind speed and the aerofoil properties, the chord, twist distributions are calculated. Then a

CFD analysis for obtaining the lift & drag coefficients (CL and CD) is done to calculate the lift

and drag forces acting on the blade.

Yangfeng Wang [28] considered two cases of turbine blades having 1m and 5m in length and

conducted damage detection technique by comparing the dynamic response analysis and mode

shape curvature methods using composite multi-layer materials. The dynamic analysis method

is used to understand the damage severity of wind turbine blades.

9
Chu and Clausen [29] used 5KW horizontal axis wind turbine blade 2.5m long to find out the

dynamic response using LXRS telemetry system which uses 3 gauges to measure the response

during operation. At small speeds, turbine operation is satisfactory, but yaw error is found to

increase as speed increased to large extent. Zheng et al. [30] found dynamic response by

considering flexible beam elements and obtained the response characteristics for a medium and

high speed wind turbine.

1.3 Scope and Objectives

Most of the literature dealt with investigation of various airfoils design, aerodynamic

evaluation by simulation tools, and the analysis of complex composite web section blades using

various techniques including FE modelling, analytical approaches as well as experimental

works. Very few works considered the power generation by blades of around 1m length for

analysis. There is a requirement of developing a user interactive software that computes the

blade data corresponding to flow conditions and generate the dynamic response using some

finite element approach.

In the present work, research is carried out with inclusion of the following aspects: Performance

of wind turbine blade models with NACA 63415 profile, airfoil blade section analysis with

taper and twist, in stationary and rotating conditions, modal data & forced vibration analysis,

using analytical and FE Modelling.

Initially, a dynamic model of the wind turbine blade is carried using Finite element modelling.

The blade design parameters are found out by using the Blade Element Momentum theory

which defines the geometry of the blade. A computer aided design model of the blade is

developed in CATIA software with wind turbine blade having lengths of 1.02 m of NACA

63415 airfoil. The model is analysed in ANSYS and the frequency (modal) data is obtained.

Further, dynamic analysis of the blades is carried out using beam finite elements and the nodal

10
forces are applied including aerodynamic lift & drag forces, centrifugal forces and gravity

loads.

1.4 Thesis Organization

The remaining thesis is organized as follows: Chapter 2 explains the dynamic

equations of motion for out of plane bending vibrations of a rotating tapered

twisted wind turbine blade, the BEM method and types of loading a wind turbine

blade is generally subjected to. Chapter 3 introduces finite element modelling on

the blade by applying centrifugal, aerodynamic, and gyroscopic loads on the blade

element. The results and discussion of the modal and dynamic analysis are

presented in Chapter 4. In Chapter 5, conclusions are made and future scope of the

work is illustrated.

11
 Chapter-2

Mathematical Modelling
2.1 Equations of Motion of a Rotating Wind Turbine Blade

Due to high slenderness ratio, often the wind-turbine blade is treated as a rotating Euler-

Bernoulli beam. Let a blade of length L is mounted on the hub of radius R at a settling angle

φs and with an inclination angle  as shown in Fig 2.1.

y

R

z
x
φs L

t

Fig 2.1 Blade configuration

Also, if  is constant angular speed, then any point on the deformed blade can be expressed

as:

 R  ( x  u ) cost  v sin  t sin  s 


 
OP   ( x  u ) sin  t  v cos t sin  s  (2.1)
 v cos s 
 

12
The kinetic and potential energies of the system are given by:

1
L  
T=    (V p ..V p )dAdx (2.2)
20A

L
1
U=   E x 2 dAdx (2.3)
20A

Where,  and E are density and elastic modulus of material, while x is normal strain given in

terms of displacements as follows

u  2 v 1  v 
2

x= r 2    (2.4)
x x 2  x 

Here, r is the distance of arbitrary point to the centroidal axis. The non-conservative virtual

work done by aerodynamic forces and moments is given by:

W  Fxx  Fyy  M zy (2.5)

The Governing partial differential equation of motion of such a beam in flap wise bending

mode can be shown as:

2  2   2 y ( x, t )  l   2 y ( x, t )
x 2  EI ( x) x 2  y ( x, t )   A( x) t 2   x f cent ( , t )d  x 2
  (2.6)
y ( x, t )
 f cent ( x, t ). 0
x

Where, fcent is the centrifugal force acting on the blade given by,

fcent  2 A( x)(r  x) (2.7)

and ρ is the density, l is the length of beam, and ξ is dummy variable for x.

By substituting fcent, we get:

EI ( x)( y IV  y IV )  2 EI ( x)( y   y )  EI ( x)( y   y )  A( x) y
(2.8)
  2   A( )(r   )d  y   A( x)(r  x) y   0
l

 x  

13
Where dot represent differential with respect to time and primes denote differential with

respect to x and A(x), I(x) are the cross section area and Polar Moment of inertia of blade

element.

2.2 Approximate Solution approach

Galerkin’s approach is used to solve the above set of equations by considering the kth order

estimate. By Galerkin’s expansion, expression for y(x,t) changes to:

k
y ( x, t )   g p ( ) p ( x) (2.9)
p 1

The boundary conditions for cantilever beam are

y(0)  0, y(0)  0, y(1)  0, y(1)  0 (2.10)

where  p (x) contains eigen functions of a stationary, Euler-Bernoulli beam satisfying the

boundary conditions (2.10) and are written as:

 p( x)  cosh  p x  cos  p x  k p (sinh  p x  sin  p x) (2.11)

cosh  p  cos i
kp  (2.12)
sinh  p  sin  p

where,  p are the eigen frequencies which are the roots of the equation 1 cos  cosh  =0.

Substituting the value of Equation (2.10) into (2.8) gives the following expression:

Re s ( y )  EI ( x) g (t ) IV ( x)  2 EI ( x) g (t ) ( x)  EI g (t ) ( x)  A( x) g(t ) ( x)

 l   (2.13)
    A( )( r   )d  g (t ) ( x)  A( x)( r  x) g (t ) ( x)  0
2

 x0  

Orthogonalizing the residual with respect to the set of  q (x) , we obtain,

l
 Re s( y)
0
q ( x)dx  0 , q=1,2,3…,k (2.14)

14
As a procedure of Galerkin’s method, the approximate discretized equations of motion for 1st

eigen frequency of λp= 1.8751 are formed and the equations in time domain are solved and the

final solution is obtained by multiplying it with mode shape function at any location x.

2.3 Loads Acting on Wind Turbine Blade

The wind turbine blade profiles are constructed by combining the concepts of 2-dimensional

airfoil and Blade Element Momentum theory. The chord and twist distributions are obtained

from BEM theory whereas as the profile can be constructed form the camber line equations

shown in Appendix I. A lofted surface is modelled in a design software to obtain the 3-

Dimensional realistic view of the model of blade. Blade Element Momentum Theory (BEM)

discretizes the whole blade into a number of elements and couples the momentum theory with

local forces acting on each blade element. The first method examines the momentum balance

of a rotating turbine in a stream tube with wind passing over it. The second method determines

the lift & drag forces. These two methods gives a series of equations to solve the blade element

cross section as independent forces are assumed to be acting on each blade element. The model

is based on Rankine Froude Momentum model which has the following assumptions: no

aerodynamic interactions amongst blade elements, velocity in direction of length of blade is

neglected, the forces are assumed to be dependent only on lift and drag coefficients.

dFt
dFx dFd
dP

Fig 2.2 Representation of Blade elements and blade geometry

15
We cannot tap the entire energy available in the wind as only the wind which passing through

the blades can be utilized in generation of power. For more power, require to have infinite

number of blades rotating with negligible thickness. Since that cannot be possible, Betz

investigated the maximum conversion efficiency and found that we can convert only 59.25%

of available wind power into electricity. This is known as the Betz Limit and it determines the

criterion we have to consider when thinking to design a wind turbine generating a specific

power. So any wind turbine blade cannot have the value of Power Coefficient C P more than

16/27. Wind turbine blade is subjected to a complex loading subjected to random wind loading.

There are basically four major types of loading any wind turbine blade undergoes.

2.3.1 Aerodynamic Loads

These loads are based on the type of airfoil profile selected. These loads are applied due to flow

of wind over the blade and are the primary loading which generates the power. The

aerodynamic loads are of three types: Thrust Force dFt, Drag Force dFd and Pitching Moment

dM. are represented in Figure 2.2.

1. Thrust Force (FT): It acts on the lower surface of the blade in perpendicular direction

as a result of the unequal pressure on the top and bottom surfaces of the blade.

1
Thrust Force: FT   a C y c eV 2 dx (2.15)
2

2. Drag Force (FD): This force acts in the plane of rotation opposite to rotation and it is

due to visocity of the fluid making the blade to lower its speed.

1
Drag Force: FD   a C z c eV 2 dx (2.16)
2

16
 3. Pitching Moment (Mp): acts along direction of the span of the blade which causes

torsion of blade.

1
Moment M p   a C m c e2V 2 dx (2.17)
2

2.3.2 Centrifugal & Gravitational Loads:

These type of loads depend of the angular Rotational speed of the blade Ω, the mass of the

blade and is expressed as:

x l
Fcent   A2 xdx (2.18)
x 0

The gravitational load is simply given as

n
FG=  me g  mblade g (2.19)
e 1

2.3.3 Gyroscopic Load

The Gyroscopic loads causes a tilt moment to occur when the blade is rotating and

simultaneously a yaw mechanism is installed for changing the blade’s orientation with respect

to the wind direction. This combined effect will produce a resulting yaw moment and tilt

moment. But in 3 bladed turbines, the yaw moment is neglected and Tilt moment Mtilt is

expressed as:

n
Mtilt= 3 k  me re2 (2.20)
e 1

17
 Chapter-3

Finite Element Model of the blade

3.1 Blade Discretization

The taper-twisted wind turbine blade is discretized into n elements using BEM

Theory where each element has length of l e , a cross sectional area of A e . A six

degree of freedom two noded Euler beam element is considered with displacement

 
vector qe  ui , vi , wi ,  i , i ,  i , u j , v j , w j ,  j ,  j ,  j T where i and j denote the node

numbers. Fig.3.1 shows the discretization of blade from root to tip.

1
2
3
4
5
6
7
8
9
10

Fig 3.1 Discretization of the blade

18
The stiffness matrix of blade element K is calculated as:

 AE  AE 
 L 0 0 0 0 0 0 0 0 0 0 
L
 12 EI z 12 EI yz  6 EI yz 6 EI z  12 EI z  12 EI yz  6 EI yz 6 EI z 
 0 0 0 0
 L3 L3
L2 L2 L3 L3
L2 L2 
 12 EI yz 12 EI y  6 EI y 6 EI yz  12 EI yz  12 EI y  6 EI y 6 EI yz 
 0 L3 L3
0 0 L3 L3
0 
 L2 L2 L2 L2 
 GJ x  GJ x 
 0 0 0
L
0 0 0 0 0
L
0 0 
  6 EI yz  6 EI y 4 EI y  4 EI yz 6 EI yz 6 EI y 2 EI y  2 EI yz 
 0 0 0 0 
 L2 L2 L L L2 L2 L L 
 6 EI z 6 EI yz  4 EI yz 4 EI z  6 EI z  6 EI yz  2 EI yz 2 EI z 
 0 0 0 0 
[ K ]e   L2 L2 L L L2 L2 L L 
  AE 0 0 0 0 0
AE
0 0 0 0 0 
 L L 
  12 EI z  12 EI yz 6 EI yz  6 EI z 12 EI z 12 EI yz 6 EI yz  6 EI z 
 0 L3 L3
0 0 L3 L3
0 
 L2 L2 L2 L2 
  12 EI yz  12 EI y 6 EI y  6 EI yz 12 EI yz 12 EI y 6 EI y  6 EI yz 
 0 L3 L3
0
L2 L2
0 L3 L3
0
L2 L2 
  GJ x GJ x 
 0 0 0 0 0 0 0 0 0 0  (3.1)
 L L 
  6 EI yz  6 EI y 2 EI y  2 EI yz 6 EI yz 6 EI y 4 EI y  4 EI yz 
 0 0 0 0 
 L2 L2 L L L2 L2 L L 
 0 6 EI z 6 EI yz  2 EI yz 2 EI z  6 EI z  6 EI yz  4 EI yz 4 EI z 
0 0 0
 L2 L2 L L L2 L2 L L 

Euler configuration is applied to each blade element to develop the equation of Kinetic energy

of blade elements. From kinetic energy of the system, the expression for element mass matrix

is given by:

1 1 
3 0 0 0 0 0 0 0 0 0 0 
6
 13 11L 9  13 L 
0 0 0 0 0 0 0 0 
 35 210 70 420 
0  11L
0 
13 9 13 L
 0 0 0 0 0 0
35 210 70 420
 Iy  Iz Iy  Iz 
0 0 0 0 0 0 0 0 0 0 
 3A 6A 
  11L L2  13 L  L2 
0 0 0 0 0 0 0 0 
 210 105 420 140 
 0 11L L2 13 L  L2 
0 0 0 0 0 0 0
[ M ]  Ae l e 
e 210 105 420 140 

1 1
 0 0 0 0 0 0 0 0 0 0 
6 3 
0 9 13 L 13  11L 
0 0 0 0 0 0 0
 70 420 35 210 
 9  13 L 13 11L 
0 0
70
0
420
0 0 0
35
0
210
0 
 Iy  Iz Iy  Iz 
0 0 0 0 0 0 0 0 0 0 
 6A 3A 
 13 L L 2
11L L2  (3.2)
0 0 0 0 0 0 0 0 
 420 140 210 105 
 0 13 L  L2  11L L2 
 0 0 0 0 0 0 0
420 140 210 105 

19
Henceforth, the linear equation system becomes

M e X e  Ce X e  K e X e  Fe (3.3)

Where Ce is internal damping matrix of the system, Fe is the generalized force vector which

contains aerodynamic, centrifugal, gravity, and gyroscopic loads which are computed using

Blade Element Momentum theory (BEM) in case of an isothermal flow of a viscous and

incompressible fluid. The expressions for thrust force dFt, drag force dFd and pitching moment

dM are given in chapter 2 and complete element force vector is defined below.

1
dFt   a C y ceVr2 dx
2
1
dFd   a C z ceVr2 dx (3.4)
2
1
dM   a C m ce2Vr2 dx
2

Where Cx, Cy, and Cm are the lift, drag and moment coefficients respectively and are given by:

C y  CL sin r  CD cosr
(3.5)
C z  CD sin r  CL cosr

Y v y

dFt
γ dFd
α
X dFx

u dP
β
x
z
dx
w
Z

Fig 3.2 Euler configuration of the blade

20
Similarly, the Centrifugal force and the gravity load vector acting on each blade element of

length dx (see Fig.3.2) in terms of elemental cross sectional area Ae given by:

dFcent  Ae  2 dx (3.6)

dP   gAe cos sin t , cos cos t , sin   dx

T
(3.7)

The total elementary force vector dFe is simplified as:

 2 l Ae 2 xi  x j   2 glAe cos sin t 

1 2 1 
 1 1 
  a lC y ceVr2  glAe cos cos t 
 4 2 
 1 1
 a lC z ceVr  glAe sin 
2 
 4 2 
 1 1 
  a lC m ce 2Vr2  l 2 I yz 
 4 2 
 0 
 0 
dFe   1 
 l Ae 2 x j  xi   glAe cos sin t 
2 1
2 2 
 1 1
 lC c V  glAe cos cos t 
2
 4 a y e r 2 
 1 1 
  a lC z ceVr2  glAe sin  
 4 2 
 1 1 
 a lC m ce 2Vr2  l 2 I yz
 4 2 
 0 
  (3.8)
 0 

The resultant solution will contain a number of ordinary differential equation. A numerical

time integration method such as Newmark method is to be employed to solve for the dynamic

responses. The steps of this method are given in Appendix III.

3.2 3D Modelling of the Blade

The values of coordinates of top and bottom surfaces of airfoil as stored in excel format are

called in commercial solid model software CATIA V5. The macro after pasting the coordinates

prompts an option either to generate only points, or points & spline, or loft the splines. The

inserted spline in the CATIA part file is either scaled to the required size or rotated to give the

21
amount of twist to the wind turbine blade. After getting the two splines, the surface modelling

feature is activated and a loft surface is formed by multi section solids option from CATIA

toolbox. After getting the loft surfaces, upper and lower surfaces are joined so that we get a

solid model ready for analysis as shown in Fig.3.3. The solid model obtained is saved in either

.parasolid or .igs format and is exported to ANSYS for further analysis as seen in Fig.3.4. The

first few natural frequencies of the system can be found out using ANSYS workbench which

will serve as a base for transient and vibrational analysis of the system. In this analysis, the

wind turbine blade is simplified to be a cantilever problem where the blade is fixed at the hub

portion.

Fig 3.3 Solid model of the twisted blade

Fig 3.4 Imported solid model

22
 The .igs file is opened in ANSYS workbench for modal analysis. The modal analysis tool

is invoked, the materials are given and the geometry is imported into the workbench. The total

deformation and first five natural frequencies are obtained and their mode shapes are recorded.

23
 Chapter-4

Results & Discussion

Dynamic analysis of such coupled fluid -structure problem required several

considerations. First step is to generate the blade geometry. Then CAD model is

formulated in 3-D, followed by its frequency analysis. 1 -D beam formulation is

then presented with the generated force system. The dynamic equations are solved

by Newmark’s time integration with zero initial conditions. Details are explained

one after the other as follows.

4.1 Blade Data Selection Procedure

An important factor to consider in generation of power of wind turbine is the blade design.

The following procedure is to be adopted in determining the blade dimensional properties.

Step 1: Determine the radius of the blade R = 1.02m

Step 2: Choose a tip speed ratio λ=6

Step 3: The number of blades B=3

Step 4: Select an airfoil: NACA 63415

NACA 63415
0.15
0.1
0.05
y/c

0
-0.05 0 0.2 0.4 0.6 0.8 1
-0.1
z/c

Fig 4.1 Aerodynamic profile of NACA 63 -415 airfoil

24
 Start

Choose an airfoil type,

λ, α, B, R

Discretize into n
elements

e = 1 to n

Find local tip speed

ratio λr

Compute φe , ce , θe

i = i+1

e>n No

Yes
Stop

Fig 4.2 Flowchart to determine blade’s geometric parameters

Step 5: Choose a chord distribution of the aerofoil

8r
c 1  cosr 
BCL (4.1)

25
 Chord Distribution
25

20
Chord c, cm
15

10

0
0 0.2 0.4 0.6 0.8 1
r/R

Fig 4.3 Chord distribution of the blade

Step 6: Discretize the blade into n elements n=10

Step 7: Choose an angle of attack α in such a way that its value is constant throughout

the blade: α=6°

Step 8: Examine the lift CL & drag coefficient CD curves for the airfoil by considering

Reynolds Number of 100,000.

Plot of CL/CD Plot of CL/α

1.5 1.5

1 1

0.5 0.5
CL
CL

0
0
-20 -10 0 10 20
0 0.05 0.1 0.15
-0.5
-0.5
-1
-1 α
CD

26
 Plot of CD/α Plot of Cm/α
0.14 0
0.12 -20 -10 -0.01 0 10 20
0.1 -0.02
0.08 -0.03
CD

Cm
0.06 -0.04
0.04 -0.05
0.02 -0.06
0
-0.07
-20 -10 0 10 20
-0.08
α α

Fig 4.4 Plots of Lift, Drag and Moment Coefficients of NACA 63415

Step 9: Obtain the lift, drag, moment coefficients for the specified angle of attack of the

blade:

Cm=-0.0597

CL=0.8765

CD=0.0258

Step 10: Obtain the angle of relative wind φr.

2 1
  tan 1  
3  e  (4.2)

Step 11: Calculate the angle of blade twist θT, from the angle of attack and angle of

relative wind.     (4.3)

27
 Twist Distribution
35

30

25
Twist (θT) ͦ

20

15

10

0
0 0.2 0.4 0.6 0.8 1
r/R

Fig 4.5 Twist distribution of the blade

4.2 Input Parameters

The parameters of the wind turbine blade considered in the present analysis are depicted in

Table 4.1. The geometrical properties of the blade elements are tabulated in Tables A1 in the

Appendix II and Inertial properties can be tabulated as shown in Table 4.2.

Table 4.1 Properties of the blade [24]

Material Glass Fiber Reinforced Plastic

Mass density of the blade 1400 kg/m3

Elastic Modulus of the blade 6 GPa

Poisson’s Ratio 0.18

No. of Elements 10

No. of Elements at the root section 1

No. of Elements in the working region 9

Length of the blade 1.02m

Radius of the hub 0.06m

28
Table 4.2 Inertial Properties of the Blade at tip-speed ratio λ=6

2
Element dr A, cm Twist θ Ixx, cm4 Iyy, cm
4
Izz, cm
4
Iyz, cm
4

1 0.12 8.365 0 11.137 5.568 5.568 0

2 0.1 42.394 29.825 934.172 196.041 738.131 339.058

3 0.1 30.701 15.891 504.374 53.747 450.626 130.524

4 0.1 20.193 10.671 217.006 13.286 203.719 38.015

5 0.1 13.745 7.379 99.700 4.206 95.493 11.849

6 0.1 9.815 5.145 50.477 1.944 48.533 5.248

7 0.1 7.311 3.539 27.863 1.071 26.792 2.661

8 0.1 5.637 2.332 16.501 0.457 16.044 0.519

9 0.1 4.470 1.393 10.349 0.278 10.071 0.159

10 0.1 3.627 0.643 6.800 0.181 6.619 0.0178

4.3 1-D Beam Finite Element Analysis Outputs

The blade under consideration has a variable cross section and 10 elements are taken, one near

the root and nine along the blade working zone. By arresting the first node, 66  66 matrices

reduce to 60 60 matrices. The computational work is carried out using MATLAB to find out

the Global Mass and Global Stiffness matrices, giving boundary conditions, and Eigen value

analysis is conducted. The following pseudo code highlights the main parts of the program.

29
Pseudo Code:

%Initialize the parameters

Ae, le, E, ρ, Iy, Iz, Iyz, J, G, ν
%Generate Global Stiffness and Global Mass Matrices
Loop:
for i=1:length(A)
Compute [K]e, [M]e;
r= (i-1)*6+1;
GKs(r:r+11,r:r+11)=K;
GK=GK+GKs;
GMs(r:r+11,r:r+11)=M;
GM=GM+GMs;
end
%Do modal reduction by imposing boundary conditions
Loop:
for i=7:66
for j=7:66
Kr (i-6,j-6)=GK (i,j);
Mr (i-6,j-6)=GM (i,j);
end
end
%Find out Eigen Values and Eigen Vectors
A1=Kr/Mr;
[y,d2]=eig(A1);
Loop:
for i=1:60
D1(i)=d2(I,i);
end
[EV,index]=sort (D1);
%Print the 1st Five Natural Frequencies
NF=sqrt(EV)/(2*pi);
fprintf(‘First five Natural Frequncies /n’);
fprintf(‘%f\n%f\n%f\n%f\n%f’, NF(1), NF(2), NF(3), NF(4), NF(5));

The MATLAB program is implemented on a Windows 8.1 operating system with 4 GB RAM

and Intel i5 2.6GHz processor. The results obtained by the eigen value analysis are tabulated

in Table 4.4.

4.3 Modal Characteristics of the blade from ANSYS 15.0

Using ANSYS 15.0 workbench, the modal analysis tool is invoked, the material properties are

given, and the geometry is imported into the workbench. Fig 4.3 shows the mesh details for 3-

30
D model using Tetrahedral beam elements. The total deformation and first five natural

frequencies are obtained and their mode shapes are recorded.

Fig 4.6 Mapped Mesh of the blade

Modal analysis is conducted and the first five natural frequencies are shown in Table 4.2 in

comparison with 1-D model. The results are found to be in close comparison by both methods.

Table 4.4 Modal Characteristics of the Blade

1-D Beam Element, 3D Solid Model,

Mode
Hz Hz

1 11.63185 11.344

2 15.39532 15.335

3 31.5512 31.202

4 77.95628 75.913

5 124.99 124.79

31
Fig. 4.7a 1st Flapwise Bending Mode shape

Fig 4.7b 1st Edgewise Bending Mode shape

Fig 4.7c Combined Bending Mode shape

Fig 4.7d 2nd Flapwise Mode shape

32
 Fig 4.7e Torsional Mode shape

Fig 4.7 Mode shapes observed in Modal analysis conducted by ANSYS

Aerodynamic and elastic forces occurring cause aero-elastic vibrations, and these vibrations

arise around all three axes. First, the rotor blades generally vibrates across the plane in which

rotor rotates in flap-wise direction (1st mode, Fig 4.7a). Further, rotor blades vibrate in the plane

edge wise (1st mode, Fig 4.7b). Further in mode shapes Fig 4.7c and Fig 4.7d, the blades

simultaneously vibrate along 2 directions (sway). The mode shape in Fig 4.7e shows additional

torsional vibration along the x axis.

4.4 Effect of Rotational Speed on Natural Frequencies

Constant speed Analysis is helpful to examine the effe ct of system parameters on

the natural frequencies of undamped free vibrations of a tapered wind turbine

blade rotating at various speeds from 0rpm to 50 rpm.

Graphs are plotted as shown in Figures 4.8 to 4.10 between the frequency of flap-

wise bending, Edge-wise bending and torsional bending modes with increase in

the rotational speed of the blade.

33
 Flapwise bending mode variation
14
11.476 11.461 11.392 11.139
12 10.325
Frequency, (Hz) 10 9.2227

0
0 10 20 30 40 50
Rotational speed Ω (rad/s)

Fig 4.8 Variation of Flap wise bending mode with respect to rotational

speed

Torsional mode variation

143 142.18
142
Frequency, (Hz)

141
140 139.47

139
137.63
138
136.56 136.76
137
136 135.36

135
0 10 20 30 40 50
Rotational speed Ω (rad/s)

Fig 4.9 Variation of Torsional bending mode with respect to rotational

speed

34
 Edgewise bending mode variation
18
15.403 14.94
16
13.804
Frequency, (Hz) 14 12.574
11.948 11.785
12
10
8
6
4
2
0
0 10 20 30 40 50
Rotational speed Ω (rad/s)

Fig 4.10 Variation of edgewise bending mode with respect to rotational

speed

It is observed from the figures 4.8 to 4.10 the value of natural frequencies are

more in torsional bending mode compared to the flap-wise bending mode and edge-

wise bending mode. It is also seen that in the torsional mode, the trend of natural

frequencies are seen to decrease until 20 rpm and then increase. This can be noted

that optimal speed of the wind turbine operation must be more than 20 rpm in order

to ensure safety of the component.

35
4.5 Effect of tip speed ratio on the chord & Twist distributions of the blade

Graphs are plotted between the distances from the root to the chord length as

shown (Figure 4.11) and twist distributions (Figure 4.12) for different values of

tip speed ratios.

30
λ=5
25 λ=6
Chord Length, c (cm)

λ=7
20

15

10

0
0.12 0.22 0.32 0.42 0.52 0.62 0.72 0.82 0.92 1.02
Distance from root, r (m)

Fig 4.11 Chord distribution variation for different taper ratios.

40
35
λ=5
Twist angle, φT (deg)

30
λ=6
25
λ=7
20
15
10
5
0
-50 . 1 2 0.22 0.32 0.42 0.52 0.62 0.72 0.82 0.92 1.02

Distance from root, r (m)

Fig 4.12 Twist distribution variation for different taper ratios

36
From the figure 4.11 it is observed that whenever the tip speed ratio (λ) increases

the chord length also increases. From the figure 4.12 it is observed that whenever

the tip speed ratio (λ) is increases the twisting angle is decreasing. This

corresponds to the fact that larger tip speed ratio blades must have lesser chord

length to ensure that they rotate faster reducing the overall centrifugal forces of

the blade.

4.6 Effect of tip speed ratio on the modal characteristics of the blade

Graph is plotted between the first five natural frequencies and mode shapes for

different tip speed ratios of 5,6 and 7 from the results of modal analysis.

160

140

120
Frequency, Hz

100

80

60 λ=5
40 λ=6
20 λ=7

0
1 1.5 2 2.5 3 3.5 4 4.5 5
Mode

Fig 4.13 Modal Variation for different taper ratios

From the figure 4.13 it is observed that lower tip speed ratios have higher values

of modal frequencies. Compared to λ=6 & 7, where the trend of modal frequencies

are gradually increasing, the frequencies are found to be increasing at gradual

mode until 4 th natural frequency for λ=5 and then slope of the curve is observed

to decrease.

37
 160

140

120
Frequency, Hz
100

80
Flapwise
60
Edgewise
40
Torsional Mode
20

0
5 5.5 6 6.5 7
Tip speed ratio λ

Fig 4.14 Variation of flap wise, edgewise and torsional modes for different

taper ratios.

For the plot drawn (Figure 4.14) to observe the pattern of flap -wise, edge-wise

and Torsional modes at different taper ratios, it is seen that the natural frequency

values for the torsional mode the freque ncy are high and the curve reaches a peak

of 136Hz at λ=6 and then decreases. The effect of edge -wise modes with increase

of taper ratio is found to be negligible and the flap -wise modes are found to

gradually decrease from 13Hz to 9Hz.

4.7 Dynamic response Analysis

Dynamic response of the blade subjected to aerodynamic, centrifugal, gravity and

gyroscopic loads is computed by using Newmark time integration algorithm

(parameters α=0.25 & δ=0.5). The MATLAB code employed is provided in the

Appendix IV.

38
The FFT spectrum of the blade dynamic response of the displacement in x direction

at the blade tip is shown in Fig 4.15.

Single sided maginitude spectrum

0.05

0.045

0.04

0.035
Tip of blade

0.03
Magnitude

0.025

0.02
FFT response at

0.015

0.01

0.005

0
10 11 12 13 14 15
Frequency HZ

Fig 4.15 FFT spectrum of blade dynamic response.

It shows that the fundamental frequency f 1 =10.46 Hz and second frequency

f2=12.7 Hz for rotational speed (Ω) of 50 rad/s. These frequencies are found to be

slightly more than the free vibration frequency due to the additional centrifugal

stiffening & geometric stiffening effect due to the rotational velocity of the blade.

39
The Edge-wise flexure angle response (β) and the f lap-wise Flexure Angle

response (γ) are shown in the Figure 4.16 and Figure 4.17.

-4
x 10 node trajectory at end of blade
0

-2
Response, β

-4
Angledisplacement

-6

-8

-10
Flexureedgewise

-12

-14

-16

-18

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

time in sec

Fig 4.16 Edge-wise beating at blade end

node trajectory at end of blade

0.012

0.01
Flexure Angle Response, γ
flapwise displacement

0.008

0.006

0.004

0.002

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time in sec

Fig 4.17 Flap-wise beating at the blade end

It is observed that amplitudes of edge -wise beating is less (in range of 1*10 -4 ) but

the frequency of vibration is more whereas in the flap-wise beating, the amplitudes

are more (in order of 0.012) but frequency of vibration is less.

40
The edge-wise and flap-wise beating of the blade subjected to only aerodynamic

loads are shown in the Fig 4.18 gives the edgewise beating and Fig 4.19.

-4
x 10 node trajectory at end of blade
0

-2

-4
edgewise displacement

-6

-8

-10

-12

-14

-16

-18

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

time in sec

Fig 4.18 Edge-wise beating at blade end subjected to only aerodynamic loads

node trajectory at end of blade

0.012

0.01
flapwise displacement

0.008

0.006

0.004

0.002

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time in sec

Fig 4.19 Flap-wise Beating at the blade end subjected to only aerodynamic

loads

It is observed that from comparison of Figures 4.16,4.18 and 4.17, 4.19 that

aerodynamic loads contribute most of the loading on the blad e. There is negligible

difference with responses of considering all loads, and aerodynamic loads alone.

But since aerodynamic load are profile dependent, and not time dependent, its

effect on dynamic displacement response is null.

41
Figures 4.20 shows dynamic response of blade in x direction & Figure 4.21 shows

the edge-wise beating response of blade subjected to only centrifugal loads.

Single sided maginitude spectrum

0.045

0.04
FFT response with only centrifugal force

0.035

0.03

0.025

0.02

0.015

0.01

0.005

0
8 9 10 11 12 13 14
Frequency HZ

Fig 4.20 FFT spectrum of blade dynamic response subjected to only

centrifugal loads.

-4
x 10 node trajectory at end of blade

-2

-4
edgewise displacement

-6

-8

-10

-12

-14

-16

-18
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time in sec

Fig 4.21 Edge-wise beating at blade end subjected to only centrifugal loads

42
The dynamic response shows a fundamental frequency of 10.43 Hz and second

fundamental frequency of 12.8 Hz. The amplitude of the vibration is lesser

compared to the blade dynamic response when subjected to al l loads (Fig 4.15)

The flap-wise response is found to be zero when subjected to only centrifugal

forces while there is no significant difference between the edgewise beating.

Figure 4.22 shows the dynamic response of blade in x direction & Figure 4.23

shows the flap-wise beating response of blade subjected to only gravity loads.

Single sided maginitude spectrum

0.05
FFT response due to gravitational load

0.04

0.03

0.02

0.01

1 2 3 4 5 6
Frequency HZ

Fig 4.22 FFT spectrum of blade dynamic response subjected to only

centrifugal loads.

43
 -4
x 10 node trajectory at end of blade

0.5

-0.5
flapwise displacement

-1

-1.5

-2

-2.5

-3
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time in sec

Fig 4.23 Flap-wise beating at blade end subjected to only centrifugal loads

From the Figure 4.22 it can be observed that a range of frequencies with small

amplitudes occur within short frequency range. The Flap -wise beating (Figure

4.23) is irregular in nature. Since acceleration due to gravity acts in the vertical y

direction, the edgewise beating doesn’t occur when the blade is subjected to only

gravity loads.

44
 Chapter-5

Conclusion
5.1 Summary

High speed rotating wind turbine blade dynamic analysis has been conducted in

this work. From the available literature, various forces acting on the blade are

accounted and the tapered-twisted aerofoil profile of the blade was generated as a

3D model. By computing element wise cross sectional details from 3D model, a

one dimensional finite element beam modelling was considered to discretize the

blade from the hub center. Also, a method proposed in literature for the blade

dimensions selection was adopted to get the optimum chord and twist angle when

the blade tip speed ratio, airfoil type & len gth of the blade are specified as inputs.

The entire work concentrates on the beam finite element modelling of the blade.

The modal and transient analysis studies are conducted using 10 beam elements

with 6 degrees of freedom per node. It was considered that the blade is fixed at

the hub rigidly with five degrees of freedom constrained. The effect of rotational

speed on variations of the natural frequencies with the system parameters are given

and it is found that with increase in speed, torsional modes var y at larger extent

compared to flap-wise and edge-wise modes. Three blade models having tip speed

ratios of λ=5, 6 and 7 are considered and the effect of tip speed ratio on the chord

& twist distributions and on the modal characteristics are studied. T he results show

that at tip-speed ratio (λ=6), the torsional frequencies are more. Dynamic response

by considering various loading conditions of blade displacement, and edge -wise,

flap-wise beating responses are obtained and studied. Results are very interesting

45
indicating the effects of aerodynamic load, centrifugal loads and gravity on the tip

response. These results are helpful for the design of blades to avoid resonant

conditions.

5.2 Future Scope

As a future scope, approximate solution methods for the continuous system of equations have

to be applied, so as to validate the result of finite element modelling. Material issues should be

introduced to know their effects on dynamic characteristics and failure prediction approaches

using polymer composite materials. Testing and analysis of blades can be done on varying

climatic conditions such as high humidity, cold regions or high temperature zones. The fatigue,

buckling analysis and localized surface roughness on the current blade models determining the

structural integrity in real-time approaches, the tower’s structural and dynamic interactions

must also be taken into account. Fluid structure interaction studies can be done for the flow of

wind around the blades and possibilities of formation of eddies and rotor wakes. The area of

improving the material characteristics by using layered composites itself is a huge research

field for the interested because it offers innumerable combinations of materials to be used to

improve effectiveness of the blade.

46
 References

1. Nitin Tenguria, N.D. Mittal, Siraj Ahmed, “Aerodynamic Design of a HAWT Blade for

Indian Wind Condition”, International Journal of Advances in Engineering, Science &

Technology (IJAEST), Vol. 1 No.7, September – November 2011, pp 37-41.

2. John McCosker, “Design and optimization of a small wind turbine”, Rensselaer

Polytechnic Institute, Hartford, Connecticut, December 2012, pp 1-56.

3. M. Anjali, C. Tara Sasanka, Ch. Deva Raj, K. Ravindra, “Meta Heuristic Method for

the Design Optimization of a Wind Turbine Blade”, International Journal of Current

Engineering and Technology, ISSN 2277 – 4106, Special Issue -3, April 2014, pp 175-

179.

4. Chalothorn Thumthae, Tawit Chitsomboon, “Optimal Pitch for untwisted blade

horizontal axis wind turbine”, 2nd Joint International Conference on “Sustainable

energy and environment SEE, November 2006, pp 1-6.

5. Juan Mendez, David Greiner, “Wind blade chord and twist angle optimization by using

genetic algorithms”, Intelligent systems & Numerical Applications in Engineering,

Spain 2011, pp 4-19.

6. Wang Yongzhi, Li Feng, Zhang Xu, Zhang Weimin, ”Composite Wind Turbine Blade

aerodynamic and structural integrated design optimization based on RBF meta-

model”, China Academy of Aerospace Aerodynamics, May 2007, 830: 10-18.

7. K. Dykes, A. Platt, Y. Guo, et al. “Effect of Tip speed constraints on the optimized

design of a wind turbine”, National renewable energy laboratory, NREL/TP-5000-

61726, October 2014, pp 1-77.

8. Grant Ingram, “Wind Turbine Blade Analysis using the Blade Element Momentum

Method, Version 1.1”, October 18, 2011, pp 1-21.

47
9. Naishadh G. Vasjailiya, Sathya N. Gangadharan, “Aero-structural design optimization

of composite Wind Turbine Blade”, 10th World Congress, 2013, pp 201-217.

10. Ahmad Sedaghat, M. El Haj Assad, Mohamed Gaith, “Aerodynamics Performance of

continuously variable speed horizontal axis wind turbine with optimal blades”, Energy

Vol 77, December 2014 752-769.

11. Scott Larwood, C.P. Van Dam, Daniel Schow, “Design Studies of swept wind turbine

blades”, Renewable Energy 71 (2014) 563 – 571.

12. K. Turgut Gursel, Tufan Coban, Aydogan Ozdamar, “Vibration Analysis of rotor

blades of a farm wind power plant”, Mathematical and Computational applications,

Vol. 17, No. 2, 2012, pp. 164-175.

13. Gunner C. Larsen, Morten H. Hansen, Andreas Baumgart, Ingemar Carlen, “Modal

Analysis of Wind Turbine Blades”, Risφ National Laboratory, Roskilde, Denmark

February 2002, pp 1-72.

14. Chao Liu, Dongxiang Jiang and Jie Chen, “Vibration Characteristics on a Wind

Turbine Rotor using Modal and Harmonic Analysis of FEM”, 978-1-4244-8921-3/10

IEEE 2010, pp 1-5.

15. O Pabut, G Allikas, H Herranen, R Talalaev, K Vene, “Model Validation and structural

analysis of a small wind turbine blade”, 8th International DAAAM Baltic Conference,

Industrial Engineering, 19-21 April 2012, pp 1-10.

16. Saad Sami, Behzad Ahmed Zai, M. Amir Khan,”Dynamic Analysis of a 5KW Wind

Turbine Blade with experimental Validation”, Journal of Space Technology, Vol. 4,

No-1, July 2014, pp 82-87.

17. Fangfang Song, Yihua Ni, Zhiqiag Tan, “Optimization Design, Modelling and Dynamic

Analysis for Composite Wind Turbine Blade”, International Workshop of Automobile,

Power and Energy Engineering, Procedia Engineering, 16 (2011) 369-375.

48
18. Zhu Jie, Cai Xin, Pan Pan, “Static and Dynamic Characteristics Study of Wind Turbine

blades”, Advanced Materials Research Vols 433-440 (2012) pp 438-443.

19. Xin Cai, Pan Pan, Jie Zhu, Rongrong Gu, “The Analysis of the aerodynamic character

and structural response of Large Scale Wind Turbine Blades”, 6, , Journal of Energy,

2013, pp 3134-3148.

20. T. Inoue, Y. Ishida and T. Kiyohara, “Non Linear Vibraiton Analysis of wind turbine

blade (occurrence of the super harmonic resonance in out of plane vibrations of the

elastic blade”, J. Vibration and acoustics, Trans. ASME, Vol. 134, pp. 0310091 -10,

2012.

21. L. Li, Y. Li, Q. Liu and H. Lv, “Dynamic Characteristics of lag vibration of wind

turbine blade”, Acta Mechanica Solids Sinica, Vol. 26, pp. 592 – 602.

22. L. Li, Y. H. Li, Q. K. Liu, H. W. Lv, “A Mathematical Model for Horizontal axis wind

turbine blades”, Applied mathematical Modelling 38 (2014) 2695-2715.

23. L. Li, Y. H. Li, H. W. Lv, Q. K. Liu, “Flapwise dynamic response of a wind turbine

blade in super-harmonic response”, Journal of Sound and Vibration 331 (2012) 4025-

4044.

24. Hamdi H, Mrad C, Hamdi A, Nasri R, “Dynamic response of a horizontal axis wind

turbine blade under aerodynamic, gravity and gyroscopic effects”, Applied Acoustics,

July 2013, pp 1-11.

25. V. Karadag, “Dynamic analysis of Practical Blades with Shear center effect”, Journal

of Sound and Vibration (1984) 92(4), 471-490.

26. Martin Nymann Svendsen, Steen Krenk, jan Hogsberg, “Resonant Vibration control of

rotating beams”, Journal of Sound and Vibration 330 (2011) 1877-1890.

49
27. M. Keerthana, M. Sriramkrishnan, T. Velayutham, A. Abraham, S. Selvi Rajan, K. M.

Parammasivam, “Aerodynamic analysis of small Horizontal axis wind turbine using

CFD”, Journal of Wind and Engineering, Vol. 9, No.2, July 2012, pp. 14-28.

28. Yangfeng Wang, Ming Liang, Jiawei Xiang, “Damage detection method for wind

turbine blades based on dynamic analysis and mode shape difference curvature

information”, Mechanical Systems and signal processing 48 (2014) 351-367.

29. Mu Chu, Philip Clausen, “The Dynamic Response of Small Horizontal Axis Wind

Turbine Blade operating in Highly Turbulent Flow”, Solar 2014, 52nd Annual

conference of the Australian Solar Council, Melbourne, May 2014, 2323 1-17.

30. Yuqiao Zheng, Rongzhen Zhao, Hong Liu, “Dynamic response of flexible wind turbine

blade”, Telkomnika, Vol 11, No. 12, December 2013, pp, 7052-7057.

50
 Appendix I

NACA Airfoil Designations

I.1 NACA 4 digit series

The NACA 4 digit series is generally denoted by NACAmpxx,

Where,

- m indicates max. value of mean line coordinates in % of chord,

- p is the leading edge distance to location of maximum camber in tens of chord

- xx indicates the thickness of the section as % of chord.

From getting the above parameters, the camber line can be expressed as

Zc m
c

 2 2 p   2
p

for 0 ≤ ξ ≤ p (A.1)

Zc

m
c 1  p 2

1  2 p  2 p   2 
for p ≤ ξ ≤ 1 (A.2)

The Thickness distribution is given by,

Zth = 5c(0.2969 ξ0.5 – 0.1260 ξ – 0.3516 ξ2 + 0.2843 ξ3 – 0.1015 ξ4) (A.3)

The coordinates of the airfoil profile can be obtained as:

X u  X  Z th sin  a
Z u  Z  Z th cos a (A.4)

X l  X  Z th sin  a
Z l  Z  Z th cos a (A.5)

51
Where (Xu, Zu) are the coordinates of upper surface of airfoil, and (Xl, Zl) are the coordinates

of lower surface of airfoil

0.15

0.1

0.05

-0.05

-0.1
-20 0 20 40 60 80 100

Fig A1 Aerodynamic profile of NACA 4412 profile

I.2 NACA 63nnn series:

NACA 63nnn series are slowly gaining importance as profiles for wind turbineas they can

operate for less speed turbines giving optimal generation of power.

For the NACA 63nnn series, the general notation is NACA6p-lxx.

Where,

- 6 indicates the series name,

- p is the distance of where pressure is minimum on chord in terms of tens of chord

length,

- The digit l indicates the design lift coefficient in tenths of chord.

- The two digits xx indicate the thickness of section as % of chord.

52
The Basic camber line equations (when axial induction factor is 0, uniform load is given over

the span of chord), are given by:

Y CL  X   X  X  X 
 1   ln 1    ln  
c 4  c  c  c  c  (A.6)

dy C Li   x  x 
 ln 1  c   ln  c 
dx 4      (A.7)

NACA 63415
0.15

0.1

0.05
y/c

0
0 0.2 0.4 0.6 0.8 1
-0.05

-0.1
z/c

Fig A2 Profile of NACA 63-415 airfoil drawn in MATLAB

The NACA 63nnn series have gained prominence as profiles of modern composite wind turbine

blades due their advantage of higher power coefficient. Thus these are replacing conventional

profiles of blades suchas NACA 4digit and 23xxx series.

53
Table A1 Coordinate data of NACA 63415 Airfoil

Coordinates of NACA 63415 profile

Upper Surface Lower Surface

Xu Zu Xl Zl

0 0 0 0

0.00525 0.01585 0.00975 -0.0131

0.02198 0.02964 -0.0222 0.02802

0.0466 0.04264 0.0534 -0.03

0.09647 0.06077 0.10353 -0.0401

0.19705 0.08279 0.20295 -0.051

0.298 0.09362 0.302 -0.0547

0.39905 0.09527 0.40095 -0.0524

0.5 0.08871 0.5 -0.0446

0.6007 0.07595 0.5993 -0.0331

0.70106 0.05877 0.69894 -0.0199

0.80102 0.039 0.79989 -0.0072

0.90059 0.01884 0.89941 0.00184

1 0 1 0

54
 Appendix II
Table A1 Geometrical Properties of Blade Elements obtained from initial program for Tip speed ratio λ=5.

φr α
r λr CL CD Cm ce θT V Cy Cz dFt dFd dM
(deg) (deg)

0.12 0.588 0.8765 0.0258 -0.0597 0.264 39.68 6 33.68 5.7161 0.5796 0.6909 0.3066 0.36556 -0.0083

0.22 1.078 0.8765 0.0258 -0.0597 0.255 28.55 6 22.55 5.9060 0.4416 0.7821 0.2414 0.42757 -0.0083

0.32 1.568 0.8765 0.0258 -0.0597 0.216 21.67 6 15.67 6.9853 0.3477 0.8240 0.2248 0.53276 -0.0083

0.42 2.058 0.8765 0.0258 -0.0597 0.181 17.27 6 11.27 8.3485 0.2848 0.8446 0.2201 0.65265 -0.0083

0.52 2.549 0.8765 0.0258 -0.0597 0.153 14.28 6 8.280 9.8394 0.2412 0.8557 0.2196 0.77935 -0.0083

0.62 3.039 0.8765 0.0258 -0.0597 0.132 12.14 6 6.141 11.398 0.2095 0.8623 0.2211 0.90976 -0.0083

0.72 3.529 0.8765 0.0258 -0.0597 0.116 10.54 6 4.546 12.999 0.1857 0.8664 0.2235 1.0424 -0.0083

0.82 4.019 0.8765 0.0258 -0.0597 0.103 9.313 6 3.313 14.625 0.1673 0.8691 0.2264 1.17648 -0.0083

0.92 4.509 0.8765 0.0258 -0.0597 0.092 8.334 6 2.334 16.269 0.1525 0.8709 0.2297 1.31154 -0.0083

1.02 5 0.8765 0.0258 -0.0597 0.084 7.539 6 1.539 17.926 0.1405 0.8723 0.2332 1.4473 -0.0083

55
Table A2 Geometrical Properties of Blade Elements obtained from initial program for Tip speed ratio λ=6.

φr α
r λr CL CD Cm ce θT V Cy Cz dFt dFd dM
(deg) (deg)

0.12 0.705 0.8765 0.0258 -0.0597 0.225 36.521 6 30.521 6.709 0.542 0.719 0.336 0.446 -0.0083

0.22 1.294 0.8765 0.0258 -0.0597 0.199 25.129 6 19.129 7.592 0.395 0.804 0.277 0.565 -0.0083

0.32 1.882 0.8765 0.0258 -0.0597 0.160 18.652 6 12.659 9.405 0.304 0.838 0.265 0.730 -0.0083

0.42 2.470 0.8765 0.0258 -0.0597 0.131 14.690 6 8.690 11.514 0.247 0.854 0.263 0.910 -0.0083

0.52 3.0588 0.8765 0.0258 -0.0597 0.109 12.06 6 6.069 13.754 0.208 0.862 0.265 1.098 -0.0083

0.62 3.647 0.8765 0.0258 -0.0597 0.094 10.22 6 4.222 16.064 0.180 0.867 0.269 1.289 -0.0083

0.72 4.235 0.8765 0.0258 -0.0597 0.082 8.856 6 2.856 18.416 0.160 0.870 0.273 1.482 -0.0083

0.82 4.823 0.8765 0.0258 -0.0597 0.072 7.808 6 1.808 20.794 0.144 0.871 0.278 1.678 -0.0083

0.92 5.411 0.8765 0.0258 -0.0597 0.065 6.979 6 0.979 23.190 0.132 0.873 0.283 1.874 -0.0083

1.02 6 0.8765 0.0258 -0.0597 0.059 6.308 6 0.308 25.599 0.121 0.874 0.288 2.070 -0.0083

56
Table A3 Geometrical Properties of Blade Elements obtained from initial program for Tip speed ratio λ=7.

φr α
r λr CL CD Cm ce θT V Cy Cz dFt dFd dM
(deg) (deg)

0.12 0.8235 0.8765 0.0258 -0.0597 0.1925 33.685 6 27.685 7.8467 0.5075 0.7436 0.3686 0.5400 -0.0083

0.22 1.5098 0.8765 0.0258 -0.0597 0.1578 22.345 6 16.345 9.5701 0.3570 0.8204 0.3163 0.7267 -0.0083

0.32 2.1960 0.8765 0.0258 -0.0597 0.1232 16.321 6 10.321 12.259 0.2710 0.8484 0.3075 0.9626 -0.0083

0.42 2.8823 0.8765 0.0258 -0.0597 0.0990 12.755 6 6.7557 15.252 0.2186 0.8605 0.3087 1.2148 -0.0083

0.52 3.5686 0.8765 0.0258 -0.0597 0.0822 10.435 6 4.4359 18.379 0.1841 0.8666 0.3132 1.4743 -0.0083

0.62 4.2549 0.8765 0.0258 -0.0597 0.0700 8.8172 6 2.8172 21.577 0.1598 0.8700 0.3192 1.7376 -0.0083

0.72 4.9411 0.8765 0.0258 -0.0597 0.0608 7.6273 6 1.6273 24.817 0.1419 0.8721 0.3259 2.0033 -0.0083

0.82 5.6274 0.8765 0.0258 -0.0597 0.0538 6.7175 6 0.7175 28.084 0.1281 0.8735 0.3331 2.2705 -0.0083

0.92 6.3137 0.8765 0.0258 -0.0597 0.0481 6.0000 6 0.3512 31.36 0.1172 0.8743 0.3405 2.5387 -0.0083

1.02 7 0.8765 0.0258 -0.0597 0.0435 5.4200 6 0 34.666 0.1084 0.8750 0.3480 2.8075 -0.0083

57
 Appendix III

Newmark Time Integration scheme

Time Integration schemes are methods to find the discrete time responses.

Newmark method of time integration belongs to discrete time interval type where

the discrete time responses are obtained by giving input of the Global Mass,

Stiffness, Damping Matrices and Force vectors as inputs for solving the time

domain response.

The Velocity is expressed as:

X  X  (1   )X   X t 

t  dt t n t n t dt (A.8)

The Displacement is expressed as:

X t dt   X t  X t t   1   Xt   Xt dt  (A.9)

 2  

The governing differential equations will be:

M e X t dt  Ce X t dt  K e X t dt  Ft dt (A.10)

The procedure to develop the Velocity and Acceleration responses is as shown

below:

Step 1:

The Mass[M] e , Stiffness[K] e and Coupling matrices[C] e are taken.

Step 2:

58
The values of initial displacement {X 0 } and velocity X 0 are given.  
Step 3:

By choosing the values of dt, α and δ n , the values of constants a 0 to a 7 are

calculated as:

1  1 1
ao  , a1  n , a 2  , a3  1 ,
 dt 2
 dt   dt  2

n dt   
a4   1 , a5   n  2  and a6  dt(1   n )
 2 

Step 4:

Form the effective stiffness matrix and effective load matrix as:

[ K eff ]  [ K ]  a0 [M ]  a1[C] (A.11)

f  f  [M ](a {X }  a {X }  a {X }  [C](a {X }  a {X }  a {X })

t  dt t 0 t 2 t 3 t 1 t 4 t 5 t

Step 5:

Solve { X t  dt }  [ K eff }1{ f t  dt }

Step 6:

The acceleration and velocities at time t+dt can be written as:

X   a X  X   a X  a X 

t  dt 0 t  dt t 2 t 3 t
(A.12)

X  X  a X  a X 

t  dt t 6 t 7 t  dt (A.13)

59
 Appendix IV
MATLAB codes of the computations done to find out the Geometric Mass,

Stiffness matrices and Global Force vector, are presented below. Then code for

conducting Newmark Integration to get dynamic response spectrum is also

explained below.

III.1 MATLAB Code to Compute Global Mass and Stiffness Matrices

%Function to compute global Mass and Stiffness Matrices

function[GMr,GKr]=Global_Mass_Stiffness215()
clc
clear all
%Initialize the parameters such as Length, Area, Young’s Modulus, density
of each blade element,
A=[8.36529 42.39464 30.70165 20.19344 13.74556 9.815455 7.311195 5.63727
4.47042 3.627475]*1e-4;
E=6*1e9;den=1400;
L=[12 10 10 10 10 10 10 10 10 10]*1e-2;
Iz=[5.56868 738.1311 450.626805 203.71994 95.493375 48.53348 26.79224
16.04464 10.07106 6.61904]*1e-8;
Iy=[5.56868 196.0412 53.74758 13.28684 4.2068 1.94449 1.071425 0.45722
0.278195 0.18105]*1e-8;
Iyz=[0 339.0584 130.5248 38.01509 11.8492 5.248325 2.661715 0.519385
0.15931 0.017845]*1e-8;
G=2.5424*1e9;
J=[11.13736 66.25 53.38538 94.18572 72.81243 55.32523 41.51267 30.71831
22.44229 16.11037 11.59907 8.17581]*1e-8;
GK=zeros(66,66);Gks=zeros(66,66);GM=zeros(66,66);GMs=zeros(66,66);

a=zeros(length(A));b=zeros(12);c=zeros(12);d=zeros(12);e=zeros(12);f=zeros(
12);g=zeros(12);h=zeros(12);i1=zeros(12);j=zeros(12);

for i=1:length(A)

a=(A(i)*E)/L(i);
b=(12*E*Iz(i))/(L(i)^3);
c=(6*E*Iz(i))/(L(i)^2);
d=(12*E*Iy(i))/(L(i)^3);
e=(6*E*Iy(i))/(L(i)^2);
f=G*J(i)/L(i);
g=4*E*Iy(i)/L(i);
h=2*E*Iy(i)/L(i);
i1=4*E*Iz(i)/L(i);
j=2*E*Iz(i)/L(i);
l=(12*E*Iyz(i))/(L(i)^3);
m=(6*E*Iyz(i))/(L(i)^2);
n=(4*E*Iyz(i))/L(i);
o=(2*E*Iyz(i))/L(i);

60
 K=[a 0 0 0 0 0 -a 0 0 0 0 0; 0 b l 0 -m c 0 -b -l 0 -m c;0 l d 0 -e m 0
-l -d 0 -e m;0 0 0 f 0 0 0 0 0 -f 0 0;0 -m -e 0 g -n 0 m e 0 h -o;0 c m 0 -
n i1 0 -c -m 0 -o j;-a 0 0 0 0 0 a 0 0 0 0 0;0 -b -l 0 m -c 0 b l 0 m -c;0
-l -d 0 e -m 0 l d 0 e -m;0 0 0 -f 0 0 0 0 0 f 0 0;0 -m -e 0 h -o 0 m e 0 g
-n;0 c m 0 -o j 0 -c -m 0 -n i1];

r=(i-1)*6+1;
Gks(r:r+11,r:r+11)=K;
GK=GK+Gks;
Gks=zeros(66,66);
end

for i=1:length(A)

a=1/3;
b=13/35;
c=(Iy(i)+Iz(i))/(3*A(i));
d=L(i)^2/105;
e=(11*L(i))/210;
f=1/6;
g=9/70;
h=(13*L(i))/420;
j=(Iy(i)+Iz(i))/(6*A(i));
l=L(i)^2/140;
area=A(i);leng=L(i);
M=[a 0 0 0 0 0 f 0 0 0 0 0;0 b 0 0 0 e 0 g 0 0 0 -h;0 0 b 0 -e 0 0 0 g
0 h 0;0 0 0 c 0 0 0 0 0 j 0 0;0 0 -e 0 d 0 0 0 -h 0 -l 0;0 e 0 0 0 d 0 h 0
0 0 -l;f 0 0 0 0 0 a 0 0 0 0 0;0 g 0 0 0 h 0 b 0 0 0 -e;0 0 g 0 -h 0 0 0 b
0 e 0;0 0 0 j 0 0 0 0 0 c 0 0;0 0 h 0 -l 0 0 0 e 0 d 0;0 h 0 0 0 -l 0 -e 0
0 0 d]*(den*area*leng);
r=(i-1)*6+1;
GMs(r:r+11,r:r+11)=M;
GM=GM+GMs;
GMs=zeros(66,66);

end

GKr=zeros(61,61);GMr=zeros(61,61);

%Gkr and Gmr are the reduced matrices after giving initial boundary
%conditions

for i=6:66
for j=6:66
GKr(i-5,j-5)=GK(i,j);
GMr(i-5,j-5)=GM(i,j);
end
end

III.2 Function to Compute Globalized Force Vector

%Function to compute the global force vector

function[GFv]=global_force215()
clear all
clc
dt=0.001;T=1;
F=zeros(12,1);GFv=zeros(61,length(dt:dt:T));

61
count=1;
for t=dt:dt:T

GFv(:,count)=force_gen215(t);
count=count+1;
end
%function to generate generalized force vector for a particular time period
function[GFr]=force_gen215(t)
A=[8.36529 42.39464 30.70165 20.19344 13.74556 9.815455 7.311195 5.63727
4.47042 3.627475]*1e-4;
L=[12 10 10 10 10 10 10 10 10 10]*1e-2;
den_a=1.225;den=1400;
ome=50;
x=[0 0.12 0.22 0.32 0.42 0.52 0.62 0.72 0.82 0.92 1.02];
g=9.81;
Fks=zeros(66,1);GFv=zeros(66,1);
Cy=[0 0.468969449 0.350179036 0.276010116 0.227869405 0.194719307
0.170689529 0.152540617 0.138378381 0.127033037];
Cz=[0 0.762117034 0.821603907 0.846550333 0.858454545 0.864843104
0.868593723 0.870949967 0.872509192 0.873583902];
c=[1.631795 21.21254432 17.98410485 14.59462041 12.05483529 10.19619802
8.805764814 7.736053205 6.891462789 6.209478139]*1e-2;
Vr=[0 6.709276501 7.592497881 9.405770927 11.51439555 13.75465549
16.06476931 18.41629808 20.79435991 23.1904253 25.59925995];
Cm=[0 -0.0597 -0.0597 -0.0597 -0.0597 -0.0597 -0.0597 -0.0597 -0.0597 -
0.0597];
Iyz=[0 339.058415 130.524795 38.01509 11.8492 5.248325 2.661715 0.519385
0.15931 0.017845]*1e-8;

for i=1:length(A)

if i==1
F=zeros(12,1);
else
ti=ome*t*pi/180;
a=(1/6)*den*L(i)*ome*ome*A(i)*(2*x(i)+x(i+1))-
0.5*den*g*L(i)*A(i)*sin(ti);
b=(1/4)*den_a*L(i)*Cy(i)*c(i)*Vr(i)*Vr(i)-
0.5*den*g*L(i)*A(i)*cos(ti);
d=(1/4)*den_a*L(i)*Cz(i)*c(i)*Vr(i)*Vr(i);
e=(1/4)*den_a*L(i)*Cm(i)*c(i)*c(i)*Vr(i)*Vr(i)-
0.5*den*L(i)*ome*ome*Iyz(i);
f=0;
h=0;
j=(1/6)*den*L(i)*ome*ome*A(i)*(x(i)+2*x(i+1))-
0.5*den*g*L(i)*A(i)*sin(ti);
k=(1/4)*den_a*L(i)*Cy(i)*c(i)*Vr(i)*Vr(i)-
0.5*den*g*L(i)*A(i)*cos(ti);
l=(1/4)*den_a*L(i)*Cz(i)*c(i)*Vr(i)*Vr(i);
m=(1/4)*den_a*L(i)*Cm(i)*c(i)*c(i)*Vr(i)*Vr(i)-
0.5*den*L(i)*ome*ome*Iyz(i);
n=0;
o=0;

F= [a;b;d;e;f;h;j;k;l;m;n;o];
end

rec=(i-1)*6+1;
Fks(rec:rec+11,1)=F;
GFv=GFv+Fks;

62
 Fks=zeros(66,1);
end

GFr=zeros(61,1);

for i=6:66
GFr(i-5,1)=GFv(i,1);
end

III.3 Program of Implementing Newmark Method to conduct Dynamic Response

Analysis

%Program to conduct Newmark Double Integration Analysis to find out

%Frequency Reponse of the Blade acted upon by various loads.
clc
clear all
dt=0.001;
T=1;
GF=zeros(61,length(dt:dt:T));
GM=zeros(61,61);
GK=zeros(61,61);F0=zeros(61,1);

[GM,GK]=Global_Mass_Stiffness215();
GF=global_force215();
F0=force_gen215(0);
X0=zeros(61,1);
X0d=zeros(61,1);
X0dd=inv(GM)*(F0-GK*X0);

delta=0.5;
alpha=0.25*(0.5+delta)^2;
c0=1/(alpha*dt*dt);
c1=delta/(alpha*dt);
c2=1/(alpha*dt);
c3=1/(2*alpha)-1;
c4=delta/alpha-1;
c5=(dt*(delta/alpha-2))/2;
c6=dt*(1-delta);
c7=delta*dt;

Keff=zeros(60,60);
Keff=GK+c0*GM;
Kinv=inv(Keff);

i=1;
X(:,1)=X0; Xd(:,1)=X0d; Xdd(:,1)=X0dd; t=0;

for t=dt:dt:T

i=i+1;
F=GF(:,i-1);
Feff=F+GM*(c0*X(:,i-1)+c2*Xd(:,i-1)+c3*Xdd(:,i-1));

X(:,i)=Kinv*Feff;
Xdd(:,i)=c0*(X(:,i)-X(:,i-1))-c2*Xd(:,i-1)-c3*Xdd(:,i-1);

63
 Xd(:,i)=Xd(:,i-1)+c6*(Xdd(:,i-1))+c7*Xdd(:,i);

F=zeros(61,1);Feff=zeros(61,1);
end

N=length(0:0.001:1);
fs=30;%62.5 samples per second
fnyquist=fs/2; %Nyquist frequency
X_mags=abs(fft(X(56,:)));
bin_vals=[0:N-1];
fax_Hz=bin_vals*fs/N;
N_2=ceil(N/2);

subplot(2,1,1)

plot(fax_Hz(1:N_2),X_mags(1:N_2))
xlabel('Frequency HZ')
ylabel('Magnitude')
title('Single sided maginitude spectrum');
axis tight

subplot(2,1,2)
plot(0:dt:T,X(58,:));
xlabel('time (in sec)')
ylabel('flexure angle response gamma')
title('Flapwise beating at the blade end')
axis tight

64